Console Output

Training and evaluating model for: Microwave
Dataset length: 62229 windows


 NILMModel(
  (conv1d): Conv1d(9, 9, kernel_size=(3,), stride=(1,), padding=(1,))
  (lstm): LSTM(9, 256, num_layers=4, batch_first=True, dropout=0.1)
  (dropout): Dropout(p=0.1, inplace=False)
  (relu): ReLU()
  (output_layer): Linear(in_features=256, out_features=1, bias=True)
)
Epoch [1/300], Train Loss: 0.000659
Validation Loss: 0.000623
Epoch [2/300], Train Loss: 0.000565
Validation Loss: 0.000519
Epoch [3/300], Train Loss: 0.000517
Validation Loss: 0.000503
Epoch [4/300], Train Loss: 0.000491
Validation Loss: 0.000442
Epoch [5/300], Train Loss: 0.000485
Validation Loss: 0.000431
Epoch [6/300], Train Loss: 0.000461
Validation Loss: 0.000459
Epoch [7/300], Train Loss: 0.000444
Validation Loss: 0.000420
Epoch [8/300], Train Loss: 0.000426
Validation Loss: 0.000415
Epoch [9/300], Train Loss: 0.000411
Validation Loss: 0.000382
Epoch [10/300], Train Loss: 0.000400
Validation Loss: 0.000396
Epoch [11/300], Train Loss: 0.000390
Validation Loss: 0.000375
Epoch [12/300], Train Loss: 0.000390
Validation Loss: 0.000361
Epoch [13/300], Train Loss: 0.000376
Validation Loss: 0.000357
Epoch [14/300], Train Loss: 0.000353
Validation Loss: 0.000327
Epoch [15/300], Train Loss: 0.000328
Validation Loss: 0.000319
Epoch [16/300], Train Loss: 0.000319
Validation Loss: 0.000308
Epoch [17/300], Train Loss: 0.000339
Validation Loss: 0.000323
Epoch [18/300], Train Loss: 0.000318
Validation Loss: 0.000325
Epoch [19/300], Train Loss: 0.000308
Validation Loss: 0.000290
Epoch [20/300], Train Loss: 0.000295
Validation Loss: 0.000288
Epoch [21/300], Train Loss: 0.000280
Validation Loss: 0.000272
Epoch [22/300], Train Loss: 0.000270
Validation Loss: 0.000260
Epoch [23/300], Train Loss: 0.000261
Validation Loss: 0.000258
Epoch [24/300], Train Loss: 0.000249
Validation Loss: 0.000250
Epoch [25/300], Train Loss: 0.000235
Validation Loss: 0.000248
Epoch [26/300], Train Loss: 0.000243
Validation Loss: 0.000248
Epoch [27/300], Train Loss: 0.000273
Validation Loss: 0.000269
Epoch [28/300], Train Loss: 0.000242
Validation Loss: 0.000229
Epoch [29/300], Train Loss: 0.000224
Validation Loss: 0.000217
Epoch [30/300], Train Loss: 0.000212
Validation Loss: 0.000227
Epoch [31/300], Train Loss: 0.000209
Validation Loss: 0.000204
Epoch [32/300], Train Loss: 0.000208
Validation Loss: 0.000208
Epoch [33/300], Train Loss: 0.000194
Validation Loss: 0.000184
Epoch [34/300], Train Loss: 0.000195
Validation Loss: 0.000192
Epoch [35/300], Train Loss: 0.000192
Validation Loss: 0.000179
Epoch [36/300], Train Loss: 0.000182
Validation Loss: 0.000172
Epoch [37/300], Train Loss: 0.000174
Validation Loss: 0.000157
Epoch [38/300], Train Loss: 0.000190
Validation Loss: 0.000168
Epoch [39/300], Train Loss: 0.000162
Validation Loss: 0.000159
Epoch [40/300], Train Loss: 0.000169
Validation Loss: 0.000205
Epoch [41/300], Train Loss: 0.000203
Validation Loss: 0.000297
Epoch [42/300], Train Loss: 0.000230
Validation Loss: 0.000211
Epoch [43/300], Train Loss: 0.000196
Validation Loss: 0.000254
Epoch [44/300], Train Loss: 0.000227
Validation Loss: 0.000180
Epoch [45/300], Train Loss: 0.000170
Validation Loss: 0.000170
Epoch [46/300], Train Loss: 0.000181
Validation Loss: 0.000202
Epoch [47/300], Train Loss: 0.000183
Validation Loss: 0.000163
Early stopping triggered

Evaluating model for: Microwave
Validation MAE: 3.940803 W
Validation MSE: 2057.797119 W²
Validation RMSE: 45.362949 W
Signal Aggregate Error (SAE): 0.013313
Normalized Disaggregation Error (NDE): 0.487900

Training and Validation Loss

Training Loss Plot

Interactive Plot